DESCRIPTION (Applicant's abstract): Disturbances in language function were among the first clinical observations reported for individuals described as schizophrenic, with selected aspects accorded preeminent theoretical status. Currently, language disturbances are among the constellation of hallmark symptoms that are diagnostic of the schizophrenia disorder. However, the linguistic conditions and psychological functions that contribute to schizophrenics' language dysfunction are not well understood and require elucidation. The focus of this revised application remains directed to the development of a theory that will explain the language processing disturbances of schizophrenia. The meaning embodied in discourse (i.e., connected language) is generally assumed to be realized through multiple layers of linguistic structures. Our model of language dysfunction in schizophrenia proposes that schizophrenic individuals are characterized by capacity limitations that influence their ability to use fully the contextual information that is conveyed through semantic and syntactic linguistic conventions. These capacity limitations exist at multiple levels, and produce dysfunctions that range from the processing of single words in small, discrete units of semantic context (lexical identification), to the processing of the logical relationships expressed in complex sentence structures. In the FIRST award project, schizophrenics showed a reduction in their ERP sensitivity to semantic context during lexical identification (ERP priming), which was only marginally improved during pharmacological treatment. A major finding was that the source of patients' compromised ERP priming was in their response to semantically associated contexts. Functions that were correlated with the accuracy of sentence comprehension included a sensitivity to word frequency information during lexical identification as well as a capacity to hold or maintain information while the semantic and syntactic parsing is accomplished. In contrast, sensitivity to type of semantic context during lexical identification (ERP priming) was not correlated with comprehension accuracy. The specific aims of the current proposal are: (1) To increase the specification of our model of language processing disturbance in schizophrenia, we will test hypotheses about the linguistic conditions and psychological functions that produce disruptions in schizophrenics' language processing. (2) To compare language processing between schizophrenic patients, family members of schizophrenic probands, and normal controls.